Daniel@0: Daniel@0: Daniel@0: Daniel@0: Netlab Reference Manual demmdn1 Daniel@0: Daniel@0: Daniel@0: Daniel@0:

demmdn1 Daniel@0:

Daniel@0:

Daniel@0: Purpose Daniel@0:

Daniel@0: Demonstrate fitting a multi-valued function using a Mixture Density Network. Daniel@0: Daniel@0:

Daniel@0: Synopsis Daniel@0:

Daniel@0: 
Daniel@0: demmdn1
Daniel@0: Daniel@0: Daniel@0:

Daniel@0: Description Daniel@0:

Daniel@0: The problem consists of one input variable Daniel@0: x and one target variable t with data generated by Daniel@0: sampling t at equal intervals and then generating target data by Daniel@0: computing t + 0.3*sin(2*pi*t) and adding Gaussian noise. A Daniel@0: Mixture Density Network with 3 centres in the mixture model is trained Daniel@0: by minimizing a negative log likelihood error function using the scaled Daniel@0: conjugate gradient optimizer. Daniel@0: Daniel@0:

The conditional means, mixing coefficients and variances are plotted Daniel@0: as a function of x, and a contour plot of the full conditional Daniel@0: density is also generated. Daniel@0: Daniel@0:

Daniel@0: See Also Daniel@0:

Daniel@0: mdn, mdnerr, mdngrad, scg
Daniel@0: Pages: Daniel@0: Index Daniel@0:
Daniel@0:

Copyright (c) Ian T Nabney (1996-9) Daniel@0: Daniel@0: Daniel@0: Daniel@0: